I mproving the reporting of randomised trials: CONSORT Statement - PowerPoint PPT Presentation

I mproving the reporting of randomised trials: CONSORT Statement and beyond Doug Altman The EQUATOR Network Centre for Statistics in Medicine, Oxford, UK C S C S 1 M M

“The whole of medicine depends on the transparent reporting of clinical trials” Drummond Rennie, JAMA 2001 2

� “This leads one to consider if it is possible, in planning a trial, in reporting the results, or in assessing the published reports of trials, to apply criteria which must be satisfied if the analysis is to be entirely acceptable…. � “A basic principle can be set up that … it is at least as important to describe the techniques employed and the conditions in which the experiment was conducted, as to give the detailed statistical analysis of results.” � “I f cases are allotted to a control group or to a treatment group … what method of random selection is used?” [Daniels M. Scientific appraisement of new drugs in tuberculosis. Am Rev Tuberc 1950;61:751-6.] 3

Daniels M. Clinical evaluation of chemotherapy in tuberculosis. Br Med Bull 1951;7:320-6.

“… editors could greatly improve the reporting of clinical trials by providing authors with a list of items that they expected to be strictly reported.” [DerSimonian R et al, NEJM 1982] 5

CONSORT 1996 6

7

8

2001 Revision of CONSORT � Major revision begun in 2000 – published in 2001 � Checklist – major revision � Also small changes to flow diagram � Short paper (“The CONSORT Statement”) – published in 3 journals � Explanatory paper (E&E) 9

10

11

Rationale for checklist items � Necessary to evaluate the study � Evidence-based, whenever possible � Minimum set of essential items 12

The “explanation and elaboration” manuscript � To enhance the use and dissemination of CONSORT � For each checklist item: examples of good reporting and explanation, with relevant empirical evidence 13

CONSORT Extensions to other trials designs � Modifications to and possibly additions to the checklist items – Possibly also modification of the flow diagram. � Extensions were planned for 6 trial designs – cluster randomised trials – non-inferiority and equivalence trials – multi-arm parallel group trials – crossover trials – factorial trials – within-person randomised trials � Also in development – N-of-1 trials

15

Reporting of adverse events in RCTs of HAART: systematic review. [Chowers et al. J Antimicrob Chemother 2009] � Only 16/ 49 trials reported AEs with no pre-selection � 67% reported only some AEs – e.g. the most frequent, if P< 0.05, or ‘selected’ AEs � “These facts obstruct our ability to choose HAART based on currently published data.” � “Authors and editors should ensure that reporting of AEs in HAART trials follows the CONSORT guidelines for reporting on harms in randomized trials.” 16

I mplementations of CONSORT � Acupuncture (STRI CTA) � Herbal medicines � Pragmatic trials � Non-pharmacological treatments 17

18

2010 Revision of CONSORT � Meeting in January 2007 � Revised checklist � Short paper (published in 9 journals) � Revised (and expanded) explanatory paper (E&E)

Major changes in 2010 � Added 3 new items – Registration, Protocol, Funding � Added several sub-items – e.g. any important changes to methods after trial commencement, with a discussion of reasons � Made some items more specific – e.g. allocation concealment mechanism, blinding � We simplified and clarified the wording throughout � NB Changes are documented in paper 20

Blinding in CONSORT 2010 � We added the specification of how blinding was done and, if relevant, a description of the similarity of interventions and procedures � We eliminated text on “how the success of blinding (masking) was assessed” – lack of empirical evidence supporting the practice – theoretical concerns about the validity of such assessment 21

Evolution of the CONSORT Statement Outcomes � CONSORT 1996 – “Primary and secondary outcome measure(s) …” � CONSORT 2001 – “Clearly defined primary and secondary outcome measures …” � CONSORT 2010 – “Completely defined prespecified primary and secondary outcome measures, including how and when they were assessed” 22

Evolution of the CONSORT Statement I nterventions � CONSORT 1996 – “Planned interventions and their timing” � CONSORT 2001 – “Precise details of the interventions intended for each group and how and when they were actually administered” � CONSORT 2010 – “The interventions for each group with sufficient details to allow replication, including how and when they were actually administered” 23

What do we need to know about treatment allocation? � Was the allocation sequence generated in an appropriately unpredictable way, e.g. by randomization [“Sequence generation”] – How was the sequence determined? � Was the act of allocating a treatment to a patient done without any knowledge of what treatment they will get? [“Allocation concealment”] – What was the mechanism of allocation? 24 24

Description of randomization in RCTs So important that CONSORT checklist has 3-4 items: Item 8a. Method used to generate the random allocation sequence Item 8b . Type of randomisation; details of any restriction (such as blocking and block size) Item 9. Mechanism used to implement the random allocation sequence (such as sequentially numbered containers), describing any steps taken to conceal the sequence until interventions were assigned Item 10. Who generated the random allocation sequence, who enrolled participants, and who assigned participants to interventions 25 25

Good (clear) reporting Sequence generation: � “Independent pharmacists dispensed either active or placebo inhalers according to a computer generated randomization list.” � ... The randomization code was developed using a computer random number generator to select random permuted blocks. The block lengths were 4, 8, and 10 varied randomly ...” 26

Clear reporting but poor methodology “Randomization was alternated every 10 patients, such that the first 10 patients were assigned to early atropine and the next 10 to the regular protocol, etc. To avoid possible bias, the last 10 were also assigned to early atropine.” [Lessick et al, Eur J Echocardiography 2000;1:257-62] 27

Clear reporting? “Patients were assigned to either the intervention or control group, by selection of a card from a pile of equal numbers of cards for each group.” [ Lancet 2002; 360: 1455–61.] 28

� “Randomization was alternated every 10 patients, such that the first 10 patients were assigned to early atropine and the next 10 to the regular protocol, etc. To avoid possible bias, the last 10 were also assigned to early atropine.” 29

Concealed allocation? � “Randomization was carried out by having prepared in advance a small box with 50 identically sized pieces of paper folded so that they could not be read. 25 had A and 25 had B written on them. The box was shaken and one of the pieces of paper was removed from the box blindly.” [Coan et al (1980) cited by van Tulder et al. Spine 1999.] 30

“They were randomised by selecting from random numbers held in sealed envelopes” “Randomisation was performed in advance with a random number table by a hospital pharmacist not involved in the study, and treatment allocations were sealed in opaque envelopes. Investigators were blind to these allocations.” 31

Comparing trial publications with protocols – sample size and analysis � Unacknowledged discrepancies between protocols and publications – sample size calculations (18/ 34 trials), – methods of handling protocol deviations (19/ 43) – missing data (39/ 49), – primary outcome analyses (25/ 42) – subgroup analyses (25/ 25) – adjusted analyses (23/ 28) � I nterim analyses were described in 13 protocols but mentioned in only five corresponding publications [Chan et al, BMJ 2008] 33

Lancet 1999; 353: 490-3 “Electronic publication of a protocol could be simply the first element in a sequence of “threaded” electronic publications, which continues with reports of the resulting research (published in sufficient detail to meet some of the criticisms of less detailed reports published in print journals), followed by deposition of the complete data set.” 34

Sharing data is not a new idea � “Experience has shown the advantage of occasionally rediscussing statistical conclusions, by starting from the same documents as their author. I have begun to think that no one ought to publish biometric results, without lodging a well-arranged … copy of his data in some place where it should be accessible, under reasonable restrictions, to those who desire to verify his work.” Galton F. Biometrika 1901 � “…the data of almost any laboratory worker, if he conscientiously describes his technique and material, have considerable value for an indefinite period.” Dunn HL. Physiol Rev 1929 35

36

www.consort-statement.org 37